Oak – The Downland gridshell

On Sussex county's far western edge The Weald and Downland Gridshell was the first timber gridshell to be completed since Otto's first Mannheim 1974 building. In the aftermath of the opening in 2001, the Downland gridshell became a cult building, and Stirling award finalist, and its architects, Edward Cullinan Architects, best-known project.

The absence of timber gridshells prior to the Weald and Downland gridshell is understandable. They are very complicated to engineer and to build. One critical challenge had been the high rate of timber lath failures and breakages, a real problem at Mannheim. A second problem has been the lack of wood technology adequate to produce laths long and strong enough to create structurally effective structures.

In the intervening quarter of a century between Mannheim and the R&D which began the Weald and Downland gridshell project, there have been parallel revolutions in glue and wood technologies. This meant the project team could realistically contemplate laths with spans that the likes of Otto could only dream of. At Weald and Downland, the team identified a Swiss glue manufacturer, Collano Adhesives whose recently developed polyurethane glue; super strong, yet environmentally sound, adhesive could be used to join shorter pieces into long laths. This glue was used to bond six pieces of timber together into the impressively long 35 to 50 metre laths. At the same time French oak was sourced, chosen for quality and cost, and cut in France, into much shorter, 6m laths, 35mm by 50 mm in section. Shipped to Britain, it was next transported to specialist timber processors, Grecon in Newcastle. There imperfections were cut out, leaving high quality sections ranging from 30Omm to 140Omm in length. These were finger jointed back together again to form the 6m 'improved timber', before being returned to Sussex to form the extra-long laths. Some have observed a material miles issue with the French oak to which some in the project team pointed to how French oak was historically part of the medieval southern English economy, an argument diminished by the further shipping to the North of England.

Mannheim 1974
Alongside this wood and glue technology, the computer revolution has also been a key influence on the evolution in engineering capacity, enabling Buro Happold, the structural engineers, to test the gridshell for weaknesses long before it was built. Buro Happold, with its long association with Bath University, worked with the structural engineering department, to begin applying off-the-shelf software modelling to explore the bending properties and behaviour of wood, hoping to ensure minimal breakages in the finger joints.

Taken together the convergence of this computer modelling research from the synergies between wood, glue and materials technology, is a graphic example of how timber design is in the midst of a transformative revolution. This convergence is enabling the realisation of radically different, yet low energy, sustainable buildings, hitherto too impossibly complex to realise.

At the same time as this modelling work was going on the project team brought in an open-minded frame carpentry business, the locally-based Green Oak Carpentry Company (or GOCC), to carry out the physical construction of the gridshell. The convergence of hi-tech with hands-on carpentry is at the heart of how the building captured the imagination of press and public alike. The notion of the hi-tech new media engineers working in concert with in-the-body carpenters, guiding them to reposition the timber to the most exacting positions, like astronauts working on an earthbound, wooden space-station, stirs the mind, and aptly illustrates the fusion of traditional craft with twenty first century hi-tech. It was a gift for the carpenters as well, enabling them to show exactly what they could do, given half a chance. This, in a building environment, which for years has been heading in precisely the opposite direction, further and further into greater pre-fabrication and away from the uses of individual skill.

Whereas prefabricated inorganic materials are built to the exact gridshell shape, the living properties of wood make it a pliable, changeable, albeit stiffer, organic fabric. At Weald and Downland, this meant the gridshell could be prepared on the ground flat as a two dimensional lattice grid mat, before being moved into its vertical position, where it changed, or deformed, into a new, springy, and provisional shape. When this happened, the wood took up all sorts of surprise and unexpected shapes, adjusting and re-adjusting, the materials organic qualities remaining inaccessible to modelling. Indeed, not knowing how the timber actually shapes out, and doing so without it being damaged in the process, turned out to be a considerable dilemma. Previous gridshells, specifically Mannheim, had been pulled or pushed up from the ground. This was where many of the breakages had happened. As an experiment and to try and counter the breakages, the Downland gridshell was constructed on a specially prepared and expensive, PERI scaffolding system, standing initially on a 7.5 m from the floor. The gridmat was constructed on the raised scaffolding, resting flat, significantly above ground level. Then section-by-section, the scaffolding was removed, and the laths slid into the diamond shell-like form, each piece cajoled into place by the carpenters, the forest of jacks tightening into position at each of the connector nodes, the points where the intricate laths crisscrossed each other.

Only once this had happened did the elaborate whirling weave of the building's surface at last become clearly visible. The diamond latticing, comprising two layers of laths was pushed and pulled, depending on their position in the grid. The outer layer was able to slide into larger lozenges, narrower and deeper over the domes, and flatter in the roofs valleys. This surface form was originally anticipated by the computer's modelling, but once the structure was up, much fine-tuning was needed from human eyes, supported by repeated and careful measurement. The lower laths were tightened in relation to their upper partners. It was here that a lot of the fine-tuning was done by eye. The carpenters knocked the laths into position, pushed out flat areas and deepened those where the building needed greater volume. Once the final shape was reached, the whole building was measured again, comparing what was in front of the team with originally modelled form. Astonishingly, the gridshell closely echoed the shape and measurements of the original digital modelling.

This carpentry ethic has extended from the large-scale gridshell, right across the building process. One example of the teamwork, involved the original plan which envisaged positioning slotted holes at the crossing points of the laths, but when this was taken to the carpenters, who quickly pointed to the difficulties to making holes and the comparative cost involved. After a round of meetings, phone conversations, emails and faxes, a new and refined design solution emerged. The result were the 'nodal connectors', which went on to be designed, a combined effort of everyone in the team, from carpenters to engineers and architects. The result is practical in construction, while performing several engineering functions within the finished building and is also visually appealing.

In a second example, some of the laths originally required rows of ten bolts in close proximity to meet the Euro Code 5 timber safety regulations, GOCC pointed out that green oak would split due to shrinkage and was likely to split, from the number of bolts. So again by returning to drawings, further computer modelling, and contributions from the carpenters, a solution was arrived at which successfully transferred the load of the lath, whilst meeting the regs and maintaining the stress integrity of the oak.

Today, completed the Weald and Downland latticed gridshell is a remarkable building to visit. It has been covered in locally sourced Western red cedar cladding, the roofing taking on something of a contemporary version of the layered cladding found on Norwegian and Russian churches. The costs are generally considered efficient, and if the environmental costs are factored in compares particularly well to high-tech buildings. Costing £1.6 million, the buildings planned life of sixty years. Given the potential, this emergent smart timber futurism should give the boosters of high eco-tech architecture some pause for thought. Although the gridshell is very much a showcase now, many are hoping is it will spur on further buildings evolving both its form and type.


A prehistory to Sussex Gridshells

Chestnut – the Weald Gridshell